Practice patterns of molecular profiling in colorectal cancer.

Abstract

641 Background: Colorectal cancer (CRC) is a genetically heterogeneous disease. Molecular profiling (MP) using next-generation sequencing is increasingly used to personalize therapy. No guidelines currently exist regarding patient selection and optimal timing. Our goal was to describe our experience at a tertiary cancer center using MP in CRC patients. Methods: This is an IRB-approved, retrospective study in patients with CRC who underwent MP between March 2007 and August 2016. Tissue samples were sent for analysis in the following MP platforms: Foundation One, Caris, and FCCC Targeted Cancer Panel (FTCP), which tests for the 50 most common mutations. Data regarding patient demographics, mutations, and clinical outcomes were analyzed. Kaplan Meier methods were used for survival analysis using SPSS. Results: We evaluated 248 patients with CRC. 60.1% were male and 80.1% were white. The median age was 59.5 years. 66.5% had colon and 33.5% had rectal CA. Initial stages: stage 1 (2.8%), stage 2 (14.1%), stage 3 (22.9%), stage 4 (59.2%). 82.2% were tested via FTCP, 8.9% via Foundation One, and 9.3% via Caris. 60.9% had the primary tumor tested. 5.2% had no mutations, 19% had 1 mutation, 28.6% had 2 mutations, 27% had 3 mutations and 20.2% had >4 mutations. The most common mutation guiding targeted therapy was KRAS (43.5%). 50% of patients had R0 resection and 19.3% went on to targeted therapy. 76.2% of resectable patients had a metastatic recurrence. 51.9% had targeted therapy for recurrence and/or stage 4 disease. The median time from diagnosis to MP was 9.9 months overall and 2.7 months for stage 4 patients. The median time from date of recurrence to MP was 7.7 months. Median length of follow up was 1.7 years. 14.9% had no evidence of disease at last follow up, 73% were alive with disease and 10.9% had died of the disease. Median overall survival was 55.8 months (CI 41.9 - 69.7). Conclusions: MP is utilized commonly in patients with stage 4 and recurrent CRC and occurs within 2.7 months and 7.7 months respectively. Further research is underway to evaluate if the information provided by MP improves outcomes in CRC, provides novel targets and will lead to increased clinical trial accrual.